R! I Choose You!: Discussing the Great Pokémon Debate

Hello everyone,

At one point in time, I really hoped to get into a project and turn it into a series of blog post. However, I just keep finding myself starting a new obsession without burning out the old one.

This time, I want to talk to you about Pokémon, more importantly, the most important decision you will ever make in your life. Which Starter are you going to pick. Like the sorting hat from the Harry Potter series, the answer to this question will be a dominate factor on who you are as a person.

Do you:

• Pick Charmander. Set the world ablaze.
• Pick Squirtle. Cause a little mischief and have a lot of fun.
• Pick Bulbasaur. The only right answer.

I want to say beforehand. I am a fan of the Grass Starters, a member of Team Bulbasaur, and a hater of the Leonardos. However, as a data analyst I am not going to skew any numbers in order to support a narrative. Although the tone of my post my reveal my love for the little green guy, I will give all starters a fair chance to prove themselves.

Also I will try to explore as many different factors as I can to evaluate the covented “Best Starter”.

What got me interested in this project?

I originally got the idea to perform an in-depth analysis on the Kanto starters from two things sources During my latest birthday party, my friends and I had decided to celebrate at a bar. While we were drinking I noticed that two of my friends had cell phone pop sockets of different starter Pokémon from the first generation. When I pointed this out I mentioned that Bulbasaur was the best(because it’s true) and to my surprise, the conversation erupted into a full blown debate.

We had a lot of fun arguing our points but it got me thinking “How can I prove that I am right and my ill informed friends are wrong”? A few weeks later, another one of my friends brought the topic up again but this time he cited an article whose headlines read “Squirtle, I (Should) Choose You! Settling a Great Pokémon Debate with Science”

Settling a debate by Science? Count me interested!!! However once I actually read the article I was pretty disappointed. Not because it said Squirtle was the best but because how it came to that conclusion.

Here is the link to the article: https://blogs.scientificamerican.com/but-not-simpler/squirtle-i-should-choose-you-settling-a-great-pokemon-debate-with-science/

To summarize the article, it basically chooses Squirtle as the best starter for a few…odd points:

1. When comparing the type advantage/disadvantage of the starters and their evolutions compared to the gym leaders and elite four(videogame bosses), The squirtle line has the least number of bad matchups. The artcle proves this by providing a heatmap style matrix of the matchups
2. Professional Speed runners will often pick Squirtle because his access to type specific move set much earlier than the other starters.

I will explain the type match ups more later but this article really bothered me. For anyone who has played the Pokémon games, you will know that in its core it is rooted in mathematics. Something that this article mostly leaves out. Even more annoying, the article even brings up stats when it addresses a different topic “Is Magikarp the worse Pokemon”?

Lets help remedy this

How do the Starter Pokémon’s Stats compare?

To explain it briefly, stat scores are numerical values that explain the characteristic of a given Pokémon. The higher the stat, the stronger the characteristic. There are 6 stat score in total.

• HP: How much damage a Pokémon can take before fainting
• Attack: How much damage a Pokémon delivers giving a physical attack
• Special Attack: How much damage a Pokémon delivers giving a special attack
• Defense: How much physical damage is reduced once a Pokémon is hit with an attack
• Special Defense: How much special damage is reduced once a Pokémon is hit with an attack
• Speed: How fast a Pokémon is, which determines when they can act in combat

Stat scores are numerical values which have linear properties so the higher a Pokémon’s Attack score is the more physical damage it can do with its attack It amazed me that this was overlooked/excluded, especially because it is brought up in the same article while talking about a different Pokémon. So let’s examine how the Pokémon compare, stat wise.

I used this dataset from Kaggle: https://www.kaggle.com/rounakbanik/pokemon

and made a subset with just the Pokémon I am interested in(The Kanto Starters).

library(dice)
library(tidyverse)# This is R's #1 tool for data manipulation
library(VapoRWave)# This is an extended theme for ggplot. It's an aesthetic  but I really dig it

From that our dataset I manually deleted every entry but the starters Bulbasaur, Charmander, and Squirtle. I also manually recreated the stat features to make selecting easier on myself

starter_stats <- Kanto_Starters %>%
  select(name,hp:speed) %>%
  gather(hp:speed,key = "stat_type", value = "stat_value") # This puts the data in "Long Format" which greatly improves its malleability
starter_stats  

Now let’s see how our starters compare stat wise

ggplot(data = starter_stats, aes(x = stat_type, y = stat_value, fill= name))+
  geom_bar(stat = "identity", position = "dodge", color = "white")+
  scale_fill_manual(values = colorRampPalette(c("springgreen3","orangered","steelblue3"))(3))+
  labs(subtitle = "1st-gen starters by their stats",
       y= "Stat Score",
       x = "Stat Type",
       fill = "Pokémon Starters",
       title = "Which Pokémon Starter is the Best?")+
  new_retro()

Here we can see that Bulbasaur has the highest stats in 3 different categories(SP Attack, SP Defense, and HP), while Charmander has the highest in 2(Attack and Speed), and poor Squirtle only takes the win on 1(Defense).

Something that I thought was ever more amazing was that Bulbasaur beats Squirtle in all but 1 category.

It appears that when it comes to stats, Bulbasaur is the best starter.

I know what some of you are thinking, “What about evolution”? Maybe Bulbasaur’s stats start better but this changes as they evolve. Let’s take a look

Evo_Stats <- ggplot(data = starter_evo_stats, aes(x = stat_type, y = stat_value, fill= name))+
  geom_bar(stat = "identity", position = "dodge", color = "white")+
  scale_fill_manual(values = colorRampPalette(c("springgreen3","orangered","steelblue3"))(3))+
  labs(subtitle = "1st-gen starters by their stats",
       y= "Stat Score",
       x = "Stat Type",
       fill = "Pokemon Starters",
       title = "Stat Score Through Various Stages of Evolution")+
  transition_states(
    Evolution_Stage,
    transition_length = 2,
    state_length = 1.5
  ) +
  enter_fade() + 
  exit_shrink() +
  ease_aes('sine-in-out')+
  new_retro()

Evo_Stats_Save <- animate(Evo_Stats, height = 800, width = 800) # This lets me save and alter the size of the animation
anim_save("Evo_Stat_Ani.gif")# This lets me save it as a .gif

In this image you can see how each starters stats grow as the achieve the next stage in evolution(represented by the pause), becoming Ivysaur then Venusaur, Charmeleon then Charizard, and finally Wartortle then Blastoise.

Everyone changes drastically but I think Charizard really gains the most. He has the highest total stat and now ties with 3 wins. He also gets a monstrous lead in the SP_Attack stat points. Things however don’t look good for the cannon turtle. It doesn’t win a single stat. Blastoise does get more balanced a theme that will continue to show itself in this series.

At least the article was right about one thing, being least worse is something to brag about.

Here you can clearly see that Bulbasaur is the best starter, something we all knew anyways But stats are just one of the few topics I am going to discuss in this series so who knows…Next I am going to talk about Type Advantages and Disadvantages

Thanks for reading, and thanks for being patient with me! Bye guys